DESCRIPTION(provided by applicant): Colorectal cancer remains the second leading cause of malignant death, and better preventive strategies are needed. Stool testing, unlike other conventional screening approaches, is noninvasive and requires no cathartic preparation. However, widely-used fecal blood tests frequently yield inaccurate results that lower screening effectiveness and raise program costs. There is a compelling biological rationale to target altered DNA exfoliated from neoplasms into stool, and multiple DNA markers would need to be targeted due to the genetic heterogeneity of colorectal neoplasia. Early data suggest that a multi-target DNA-based assay has potential to detect screen-relevant colorectal neoplasia (cancer and advanced adenomas) with substantially higher sensitivity and specificity than that of fecal blood tests. This is a application to continue the current UO1 CA 89389 study for an additional year to prospectively assess the fecal DNA-based test as a promising new approach to the general screen-detection of colorectal neoplasia. A 4-year (approved 3 years + requested additional 1 year) cross-sectional multicenter study is underway to compare the validity of the DNA-based test and the most commonly used fecal blood test (Hemoccult) for identification of screen-relevant colorectal neoplasia in 4434 demographically representative average-risk persons using colonoscopy as a gold standard. The performance of the DNA-based test will also be compared to a surrogate for flexible sigmoidoscopy (distal 60 cm of colonoscopy) and to the combination of sigmoidoscopy + Hemoccult. The design further allows assessment of the impact of dietary, medication, demographic, pathology-related, and other covariates on test outcomes. Assays are performed blindly at central laboratories. A specimen bank (frozen stools and plasma) is maintained as an important resource for the economical evaluation of additional or future markers. A state-of-the-art web-based data management system is employed to efficiently enter and transfer data across the 29 participating centers with the highest quality control. Although recruitment for the study has successfully met the originally projected enrollment rate, the interim analysis revealed a lower-than-expected prevalence of screen relevant neoplasia. Accordingly, the additional year is necessary to provide a sample size with sufficient power to test the primary hypothesis. If the DNA-based test proves to have greater screening accuracy than fecal blood testing, this could translate into more effective cancer control and more efficient use of our limited health care resources.